AI success stories are everywhere—but dig deeper, and you’ll often find a more complex reality. Models may be deployed, but not used. Accuracy may be high, yet impact is low. The disconnect? Adoption.

At Everforth Apex, we’ve learned that AI’s value isn’t in what it can do—it’s in whether people actually use it. And that’s where many organizations fall short.  It’s not the failing of motivation, or the efficacy of the tooling, it’s the user experience that drives results.

The Model Is Not the Outcome

The 2025 Stanford AI Index cites record-breaking growth in AI deployment across industries, reporting the use of generative AI in at least one business function reached 71% in 2024, more than doubling from 33% in 2023. But here’s the uncomfortable truth: model performance metrics aren’t translating into business performance metrics. Why? Because AI isn't delivering impact until it’s integrated into the decisions, workflows, and behaviors of real people.

Too often, companies measure success by the model’s precision or processing power. What’s missing is the metric that matters most: behavior change.  People don’t change because they need to, or even sometimes when they want to.  Change happens when the underlying paradigms are met, when needs are anticipated, and when the why is intrinsic to the message.  It’s not sufficient to know what is changing, rather we must go further and express the delta from as is and to be as a natural result instead of an aspiration.

There’s a common pattern behind low adoption: 

• Human Behavior Paradigms: Humans have a natural aversion to ambiguity, resistance to change, and known patterns of consumer behavior which are not typically considered.
• Trust Gaps: Users don’t understand how the model works or what to do with its output.
• Group Thinking: Herd behavior is real. If even a few are wary of AI or convinced it can’t or won’t work, there is a danger that your efforts will be in vain.
• Lack of Enablement: Insufficient training, no champions, no support in the moments that matter.
• Poor User Experience: Clunky interfaces or recommendations that feel out of sync with how work actually gets done, which often times is as simple as considering how data gets consumed.
• Misaligned Problem-Solving: Models built for the wrong questions, or in ways that don’t fit operational context.

These aren’t technical issues. They’re design, strategy, and organizational change management issues. They are solved with a human-centered approach, one that maximizes communication of why the change is necessary.  Strategy is inherent to our everyday lives, and yet the best methods are ones that are the least obvious. The ubiquity of technology driving our ways of working, communicating, and existing measures more of our appetite for newness than our ability to adapt to it. Driving adoption is about understanding human aversions, fears, and mindsets. Creating a space for trial and error, while also guiding the process, is needed for successful long-term enablement. Usage over time is driven both by qualitative and quantitative consumption of information that points towards the desired outcomes. We need to have strategy that builds momentum and guides through the change until it becomes a steady state.

Building AI That Gets Used

That’s why our approach starts at the beginning—with partnership between our AI Solutions team and our Business Strategy & Transformation group. We ensure every AI investment is designed for the people who will use it, not just the infrastructure that will host it.

We apply human-centered design, run user pilots, and incorporate feedback loops long before deployment. We work with change managers and line-of-business leaders to embed AI in the natural flow of work, not bolt it on after the fact.  Our methodology is interwoven, and as such we are able to pivot quicker, and ultimately go further faster, by keeping enablement at the forefront and not an afterthought.

Because adoption isn’t a bonus—it’s the benchmark.

Delivering ROI, Not Just Models

True ROI isn’t about deploying AI. It’s about activating it. And that requires alignment—between the technical architecture, the business process, and the human behavior it’s meant to improve. When we help clients bridge that gap, we don’t just deliver AI projects. We deliver capability, confidence, and transformation. If your AI investments aren’t translating into real-world impact, the problem may not be the tech—it may be how it’s being adopted. 

Melissa Manion, Director, Strategy and Transformation Solutions, also contributed to this article.